Imagens da anatomia interna são essenciais para as práticas médicas. Estabelecer correlação entre elas, é um importante procedimento para diagnóstico e tratamento. Nessa dissertação, é proposta uma abordagem para correlacionar dados multidimensionais de mesma modalidade de aquisição baseando-se somente nas informações de intensidade de pixels e voxels. O trabalho foi dividido em duas fases de implementação. Na primeira, foi explorado o problema de similaridade entre imagens médicas usando a perspectiva de análise de qualidade de imagem. Isso levou ao desenvolvimento de uma técnica de dois passos que estabelece um equilíbrio entre a velocidade de processamento e precisão de duas abordagens conhecidas. Avaliou-se a qualidade e aplicabilidade do algoritmo e, na segunda fase, o método foi estendido para analisar similaridade e encontrar a localização de uma imagem arbitrária (2D) em um volume (3D). A solução minimiza o número virtualmente infinito de possíveis orientações transversais e usa otimizações para reduzir a carga de trabalho e entregar resultados precisos. Uma visualização tridimensional volumétrica funde o volume (3D) com a imagem (2D) estabelecendo uma correspondência entre os dados. Uma análise experimental demonstrou que, apesar da complexidade computacional do algoritmo, o uso de amostragem, tanto na imagem quanto no volume, permite alcançar um bom equilíbrio entre desempenho e precisão, mesmo quando realizada com conjuntos de dados de baixa intensidade de gradiente. / Images of the inner anatomy are essential for clinical practice. To establish a correlation between them is an important procedure for diagnosis and treatment. In this thesis, we propose an approach to correlate within-modality 2D and 3D data from ordinary acquisition protocols based solely on the pixel/voxel information. The work was divided into two development phases. First, we explored the similarity problem between medical images using the perspective of image quality assessment. It led to the development of a 2-step technique that settles the compromise between processing speed and precision of two known approaches. We evaluated the quality and applicability of the 2-step and, in the second phase, we extended the method to use similarity analysis to, given an arbitrary slice image (2D), find the location of this slice within the volume data (3D). The solution minimizes the virtually infinite number of possible cross section orientations and uses optimizations to reduce the computational workload and output accurate results. The matching is displayed in a volumetric three-dimensional visualization fusing the 3D with the 2D. An experimental analysis demonstrated that despite the computational complexity of the algorithm, the use of severe data sampling allows achieving a great compromise between performance and accuracy even when performed with low gradient intensity datasets.

Imagens da anatomia interna são essenciais para as práticas médicas. Estabelecer correlação entre elas, é um importante procedimento para diagnóstico e tratamento. Nessa dissertação, é proposta uma abordagem para correlacionar dados multidimensionais de mesma modalidade de aquisição baseando-se somente nas informações de intensidade de pixels e voxels. O trabalho foi dividido em duas fases de implementação. Na primeira, foi explorado o problema de similaridade entre imagens médicas usando a perspectiva de análise de qualidade de imagem. Isso levou ao desenvolvimento de uma técnica de dois passos que estabelece um equilíbrio entre a velocidade de processamento e precisão de duas abordagens conhecidas. Avaliou-se a qualidade e aplicabilidade do algoritmo e, na segunda fase, o método foi estendido para analisar similaridade e encontrar a localização de uma imagem arbitrária (2D) em um volume (3D). A solução minimiza o número virtualmente infinito de possíveis orientações transversais e usa otimizações para reduzir a carga de trabalho e entregar resultados precisos. Uma visualização tridimensional volumétrica funde o volume (3D) com a imagem (2D) estabelecendo uma correspondência entre os dados. Uma análise experimental demonstrou que, apesar da complexidade computacional do algoritmo, o uso de amostragem, tanto na imagem quanto no volume, permite alcançar um bom equilíbrio entre desempenho e precisão, mesmo quando realizada com conjuntos de dados de baixa intensidade de gradiente. / Images of the inner anatomy are essential for clinical practice. To establish a correlation between them is an important procedure for diagnosis and treatment. In this thesis, we propose an approach to correlate within-modality 2D and 3D data from ordinary acquisition protocols based solely on the pixel/voxel information. The work was divided into two development phases. First, we explored the similarity problem between medical images using the perspective of image quality assessment. It led to the development of a 2-step technique that settles the compromise between processing speed and precision of two known approaches. We evaluated the quality and applicability of the 2-step and, in the second phase, we extended the method to use similarity analysis to, given an arbitrary slice image (2D), find the location of this slice within the volume data (3D). The solution minimizes the virtually infinite number of possible cross section orientations and uses optimizations to reduce the computational workload and output accurate results. The matching is displayed in a volumetric three-dimensional visualization fusing the 3D with the 2D. An experimental analysis demonstrated that despite the computational complexity of the algorithm, the use of severe data sampling allows achieving a great compromise between performance and accuracy even when performed with low gradient intensity datasets.

Imagens da anatomia interna são essenciais para as práticas médicas. Estabelecer correlação entre elas, é um importante procedimento para diagnóstico e tratamento. Nessa dissertação, é proposta uma abordagem para correlacionar dados multidimensionais de mesma modalidade de aquisição baseando-se somente nas informações de intensidade de pixels e voxels. O trabalho foi dividido em duas fases de implementação. Na primeira, foi explorado o problema de similaridade entre imagens médicas usando a perspectiva de análise de qualidade de imagem. Isso levou ao desenvolvimento de uma técnica de dois passos que estabelece um equilíbrio entre a velocidade de processamento e precisão de duas abordagens conhecidas. Avaliou-se a qualidade e aplicabilidade do algoritmo e, na segunda fase, o método foi estendido para analisar similaridade e encontrar a localização de uma imagem arbitrária (2D) em um volume (3D). A solução minimiza o número virtualmente infinito de possíveis orientações transversais e usa otimizações para reduzir a carga de trabalho e entregar resultados precisos. Uma visualização tridimensional volumétrica funde o volume (3D) com a imagem (2D) estabelecendo uma correspondência entre os dados. Uma análise experimental demonstrou que, apesar da complexidade computacional do algoritmo, o uso de amostragem, tanto na imagem quanto no volume, permite alcançar um bom equilíbrio entre desempenho e precisão, mesmo quando realizada com conjuntos de dados de baixa intensidade de gradiente. / Images of the inner anatomy are essential for clinical practice. To establish a correlation between them is an important procedure for diagnosis and treatment. In this thesis, we propose an approach to correlate within-modality 2D and 3D data from ordinary acquisition protocols based solely on the pixel/voxel information. The work was divided into two development phases. First, we explored the similarity problem between medical images using the perspective of image quality assessment. It led to the development of a 2-step technique that settles the compromise between processing speed and precision of two known approaches. We evaluated the quality and applicability of the 2-step and, in the second phase, we extended the method to use similarity analysis to, given an arbitrary slice image (2D), find the location of this slice within the volume data (3D). The solution minimizes the virtually infinite number of possible cross section orientations and uses optimizations to reduce the computational workload and output accurate results. The matching is displayed in a volumetric three-dimensional visualization fusing the 3D with the 2D. An experimental analysis demonstrated that despite the computational complexity of the algorithm, the use of severe data sampling allows achieving a great compromise between performance and accuracy even when performed with low gradient intensity datasets.

Imaging Time-of-Flight secondary ion mass spectrometry (ToF-SIMS) has been developed to perform 2D imaging and depth profiling of biological systems with micron or submicron scale lateral resolution, which can be attributed to the advent of polyatomic ion beam particularly C60+ and new concept of ToF-SIMS instrument, the J105 3D Chemical Imager (J105). These recent advances in ToF-SIMS have opened a new dimension for biological analysis. In this study, 2D and 3D imaging have been performed on two biological systems, Xenopus laevis (X. laevis) zygote/embryo and murine embryonic fibroblasts NIH 3T3 BXB-ER cells to explore the capability of ToF-SIMS to handle the biological samples with extreme topography and high resolution depth profiling of microdomains, which still represent major challenges for the ToF-SIMS. The study on X. laevis embryo explored the capability of ToF-SIMS to handle spherical samples (approx. 1-1.2 mm in diameter), identify lipid species in mixtures of lipid extraction from the zygotes and image of an intact embryo in 2D/3D during dynamic biological events, e.g., fertilisation and early embryo development. For the first time the J105 and conventional BioToF-SIMS instrument were employed for the study of developmental biology. The major classes of lipid were identified through multiple lipid assay in a single analytical run using ToF-SIMS. Topography effects of the embryo were assessed through imaging a single intact zygote/embryo that revealed secondary ions loss at the edge of the single cell. However, the topography effects on the mass resolution could be minimised using the J105. Moreover, in situ lipid profiling of the zygote revealed different lipid compositions and intensities on the membrane of the animal and vegetal hemispheres. Furthermore, high resolution imaging and depth profiling that performed on a single intact cell in a time course study visualised the egg-sperm fusion sites on the membrane of the zygote 10 min post-insemination and lipids arrangement on the membrane of the embryo through the early development stages. Subcellular signalling upon the fertilisation was also spatially located on the serial cryosections of a single zygote. With the NIH 3T3 BXB-ER cells, the study firstly adopted a finely focused C60+ beam to track morphological changes and rearrangement of subcellular organelle mitochondria (0.5-2 µm) in response to the activation of Raf/ERK (extracellular signal regulated kinase) pathway using the J105. The SIMS images of the unlabelled cells showed the shifting of membrane distribution and nuclei shrinking following Raf/ERK activation. The mitochondria fluorescence probe within the cells were located 3-dimensionally using confocal microscopy and ToF-SIMS, which revealed the distribution pattern of condensing in the two sides of the nuclei following the Raf/ERK activation. Coupled with scanning electron microscopy (SEM), the three imaging modes showed good agreement in cellular morphological changes and subcellular mitochondrial rearrangement without or following Raf/ERK activation, demonstrating an integrated approaching to study the biological processes at subcellular dimension.

The rapid technological advancement and the increased awareness about the environment have introduced new digital tools to facilitate the design process in the fashion industry in which 3D technology was incorporated. Very few research has addressed the experiences of fashion designers when using 3D tools in their design process. Most of the previous research in this field focused on the technical aspects of using 3D design tools. Hence, this paper aims to investigate the experiences of fashion designers working with 3D design tools by answering the research question "What are the pros and cons of implementing 3D digital technology in the fashion industry on the design process of designers?”. Semi-structured interviews were undertaken through Zoom with eight fashion designers located in different countries, and the data obtained were analyzed through thematic analysis. Most of the 3D designers participating in this study use CLO3D as a main program to design digital garments as it is efficient and holds a high degree of control and advanced settings. The results revealed seven themes, namely: flexibility and realism, improved communication, sustainability, gap between the digital and physical prototype, necessity of proper knowledge and training, limited creativity, and customization, and lastly, lack of social acceptance. The first three are considered pros whereas the latter four are cons. The discussion of results differentiated between two stages which are sketching and prototyping where most advantages are found in the sketching phase. Therefore, further technological advancements are needed to improve the properties of virtual garments in addition to a change of stakeholder mentality in order to embrace and accept digital technology. Hence, this study will benefit later research in the field to improve 3D tools and close the gap between physical and virtual prototypes.

Les maladies ostéoarticulaires constituent une part importante des pathologies qui touchent l'appareil locomoteur humain et elles ont un impact majeur quant à l'individu et à la société. Dans ce cadre les études cinématiques présentent un outil permettant aux cliniciens une évaluation de l'état des articulations. Ces études passent par une estimation du mouvement des structures osseuses. Plusieurs approches sont actuellement utilisées dans cet objectif : passant par le suivi de marqueurs externes (fixés sur la peau), internes (fixés sur les os), la palpation, et le recalage d'images médicales (radiographies, CT, IRM). Cette dernière approche a été retenue en proposant un recalage 2D 3D pour l'estimation des positions des structures osseuses. Dans une première partie, on tire profit de la particularité d'une nouvelle modalité radiologique bi-planes basse dose (EOS) pour proposer une méthode originale et particulièrement bien adaptée et qui utilise le théorème classique de la coupe centrale de Fourier. Pour cela, dans un premier temps, au travers d'un état de l'art des approches classiques de l'estimation de mouvements, nous réalisons une classification originale multi critères (invasivité, précision, généricité) afin de bien situer les apports de cette nouvelle modalité. Dans un deuxième temps, nous proposons une alternative au schéma classique de recalage 2D 3D qui présente plusieurs avantages. Le recalage effectué dans le domaine fréquentiel établit un lien élégant entre les données 2D et 3D ce qui permet d'éviter la simulation des radiographies. La recherche des deux composantes du mouvement rigide, rotation et translation, est effectuée d'une manière séquentielle. D'abord, la rotation est déterminée par une optimisation itérative d'une métrique dépendant des modules des données. La translation se déduit, alors, directement en utilisant les phases. Une deuxième partie est consacré à la mise en oeuvre pratique de la méthode et à l'évaluation de l'implémentation sur des données de plus en plus complexes allant des données synthétiques, des fantômes puis des données EOS réelles composées de radiographies frontales et sagittales et d'une seule reconstruction 3D. Ceci permet de valider la méthode dans des conditions réelles d'acquisition. L'application de notre méthode de recalage pour l'estimation de mouvement de différentes structures : fémur, tibia, prothèses de genou et de hanche, nous a permis d'illustrer la généricité de l'approche proposée.

La démocratisation des terminaux mobiles telle que les téléphones cellulaires, les PDAs et les tablettes PC a rendu possible le déploiement de la réalité augmentée dans des environnements en extérieur à grande échelle. Cependant, afin de mettre en oeuvre de tels systèmes, différentes problématiques doivent êtres traitées. Parmi elle, la localisation représente l?une des plus importantes. En effet, l?estimation de la position et de l?orientation (appelée pose) du point de vue (de la caméra ou de l?utilisateur) permet de recaler les objets virtuels sur les parties observées de la scène réelle. Dans nos travaux de thèse, nous présentons un système de localisation original destiné à des environnements à grande échelle qui utilise une approche basée vision sans marqueur pour l?estimation de la pose de la caméra. Cette approche se base sur des points caractéristiques naturels extraits des images. Etant donné que ce type d?approche est sensible aux variations de luminosité, aux occultations et aux mouvements brusques de la caméra, qui sont susceptibles de survenir dans l?environnement extérieur, nous utilisons deux autres types de capteurs afin d?assister le processus de vision. Dans nos travaux, nous voulons démontrer la faisabilité d?un schéma de suppléance dans des environnements extérieurs à large échelle. Le but est de fournir un système palliatif à la vision en cas de défaillance permettant également de réinitialiser le système de vision en cas de besoin. Le système de localisation vise à être autonome et adaptable aux différentes situations rencontrées.

Pour comprendre et diagnostiquer des pathologies telles que la scoliose qui affectent l'organisation spatiale de notre squelette, il est essentiel d'aborder ces problématiques en trois dimensions. Dans le cadre de leur collaboration franco-canadienne, le Laboratoire de Biomécanique et le Laboratoire de recherche en Imagerie et Orthopédie ont développé des méthodes de reconstruction 3D du squelette à partir de radiographies biplanes, notamment à partir du système de radiographie base dose EOS (Biospace Med, Paris). Ces techniques permettent une analyse clinique globale du patient, en position debout et avec très peu d'irradiations. Néanmoins, le temps de reconstruction reste contraignant pour une utilisation en routine clinique. L'objectif de cette thèse est donc de progresser dans l'automatisation des méthodes de reconstruction 3D à partir de radiographies biplanes. Les méthodes développées seront appliquées au rachis thoracique et lombaire, dans le contexte spécifique de l'étude de la scoliose. Une méthode de reconstruction s'appuyant sur une description paramétrée du rachis et sur des inférences statistiques longitudinales et transversales a été proposée et évaluée. Cette méthode permet, à partir de la saisie opérateur de quelques repères anatomiques dans les radiographies, d'obtenir très rapidement (2min 30s) une reconstruction 3D pré-personnalisée du rachis ainsi que des paramètres cliniques dédiés au diagnostic de la scoliose. Une reconstruction 3D plus précise peut être obtenue en un temps relativement réduit (10min) à partir d'ajustement opérateurs du modèle, qui s'auto-améliore par inférences au fur et à mesure des retouches. Afin de poursuivre la semi-automatisation de cette méthode, des techniques de recalage 2D/3D par traitement d'image, basées sur la segmentation des radiographies mais également sur des mesures de similarités entre des radiographies simulées et les clichés réels ont été proposées. Ces algorithmes s'appuient sur des modèles pseudo-volumiques de vertèbres, plus réalistes que les modèles surfaciques couramment utilisés. Les techniques de recalage ont été intégrées dans le protocole de reconstruction utilisant une description paramétrée du rachis et des inférences, pour proposer et évaluer une nouvelle méthode de reconstruction. Ce travail de thèse ouvre des perspectives concrètes en termes d'utilisation de telles méthodes en routine clinique et permet de poser des bases importantes pour automatiser les méthodes de reconstruction 3D à partir de radiographies biplanes de l'ensemble du squelette.

The study area, Dhurnal oil field, is located 74 km southwest of Islamabad in the Potwar basin of Pakistan. Discovered in March 1984, the field was developed with four producing wells and three water injection wells. Three main limestone reservoirs of Eocene and Paleocene ages are present in this field. These limestone reservoirs are tectonically fractured and all the production is derived from these fractures. The overlying claystone formation of Miocene age provides vertical and lateral seal to the Paleocene and Permian carbonates. The field started production in May 1984, reaching a maximum rate of 19370 BOPD in November 1989. Currently Dhurnal‐1 (D-1) and Dhurnal‐6 (D-6) wells are producing 135 BOPD and 0.65 MMCF/D gas. The field has depleted after producing over 50 million Bbls of oil and 130 BCF of gas from naturally fractured low energy shelf carbonates of the Eocene, Paleocene and Permian reservoirs. Preliminary geological and geophysical data evaluation of Dhurnal field revealed the presence of an up-dip anticlinal structure between D-1 and D-6 wells, seen on new 2003 reprocessed data. However, this structural impression is not observed on old 1987 processed data. The aim of this research is to compare and evaluate old and new reprocessed data in order to identify possible factors affecting the structural configuration. For this purpose, a detailed interpretation of old and new reprocessed data is carried out and results clearly demonstrate that structural compartmentalization exists in Dhurnal field (based on 2003 data). Therefore, to further analyse the available data sets, processing sequences pertaining to both vintages have been examined. After great effort and detailed investigation, it is concluded that the major parameter giving rise to this data discrepancy is the velocity analysis done with different gridding intervals. The detailed and dense velocity analysis carried out on the data in 2003 was able to image the subtle anticlinal feature, which was missed on the 1987 processed seismic data due to sparse gridding. In addition to this, about 105 sq.km 3D seismic data recently (2009) acquired by Ocean Pakistan Limited (OPL) is also interpreted in this project to gain greater confidence on the results. The 3D geophysical interpretation confirmed the findings and aided in accurately mapping the remaining hydrocarbon potential of Dhurnal field.

La généralisation des charges non linéaires dans l'industrie cause de nombreuses perturbations dans les réseaux électriques. Ainsi, les courants harmoniques augmentent les pertes dans les transformateurs, par le biais des effets pelliculaire et de proximité. Dès lors, ceux-ci sont assujettis à un sur échauffement. Afin d'éviter une défaillance diélectrique ainsi qu'un vieillissement prématuré, la température du point chaud ne doit pas excéder celle spécifiée par la classe. Néanmoins, pour une charge donnée il est difficile de prévoir le comportement électromagnétique et thermique d'un transformateur. L'objectif de cette thèse est donc de caractériser ce fonctionnement pour se prémunir lors de la phase de conception des nuisances ultérieures. <br /><br />Une première modélisation s'appuie sur une méthode analytique et axisymétrique permettant le calcul de la distribution des densités de courant dans les enroulements. Puis, après une homogénéisation des conducteurs, une méthode de résolution par éléments finis aboutit à l'obtention de la température en régime permanent. Enfin, la méthode est vérifiée sur un transformateur de type sec et de puissance modérée, associé à un banc expérimental innovant. <br /><br />Un second modèle propose une nouvelle approche analytique et numérique. Tout d'abord, une méthode des éléments finis 3D (MEF) permet le calcul du champ magnétique. Ce calcul non linéaire est alors associé aux équations de circuit. Puis, une MEF thermique 3D fournie la distribution de température. Finalement, le modèle couplé est appliqué à un transformateur de distribution de petite puissance et instrumenté pour vérifier le suréchauffement engendré par des courants non sinusoïdaux.